We have systematically tested various protocols optimized for DNA isolation at room temperature for the efficiency of PCR amplification while omitting the inhibitory component EDTA. Combinations of DNA isolation protocols from Edwards et al. 9 and Walbot and Warren 10 were thus compared using ~2.5 mg (~3 mm²) of A. thaliana leaf tissue in 50 μl of extraction mixture, from which 1 μl was used as template in a routine PCR application. We have found that one of the variant extraction buffers, hereafter called Sucrose Solution, exhibited no change in the efficiency of PCR amplification judged by ethidium bromide-stained agarose gels in response to varying the pH from 7.0 to 8.0, or changing the salt and sucrose concentrations between 200 to 400 mM NaCl and 300 to 440 mM sucrose, respectively (data not shown). Due to the presence of sucrose in the extraction buffer, we named this nucleic acid extraction method the Sucrose Prep. In agreement with Thomson and Henry 11, we found that heating the crude extract for 10 min followed by centrifugation for 5 sec at 6000 g eliminated nearly all debris that interferes with PCR amplification.

The Sucrose Prep protocol was further optimized by controlling sampling size. Optimal results were obtained by harvesting of leaf tissue with a 500 μl Eppendorf cap punch (yielding about 10 mg fresh tissue) in 200 μl of Sucrose Solution. Multiple samples were stored on ice prior to extraction and were ground using either sterile pipette tips or plastic pestles. The extracts could be stored at 4°C overnight or at -20°C for longer term storage, but were normally used immediately. We could store samples at -20°C for up to 4 weeks with no detectable decrease in PCR amplification efficiency. Following long term storage however, samples required re-heating before PCR amplification. Successful PCR amplification was obtained with A. thaliana DNA samples extracted from 2 to 4 week-old leaf material, petals, sepals, stigmas, styles, anthers, and embryos. The maximum size of products obtained in routine PCR amplification reactions was around 3 kb. Using optimized thermal cycling conditions, no difference was observed for products up to 4 kb when compared to DNA templates prepared according to Edwards et al. 9. By limiting the amount of plant debris carried over into the mixture, PCR product sizes of up to 5.5 kb were obtained (Figure 1).

The availability of the entire genome sequence of A. thaliana ecotype Col-0 and the high abundance of genomic sequence from Landsberg erecta (Ler) allowed us to design a new set of markers that show polymorphism between these ecotypes 1213. A novel set of 51 DNA markers were identified at regular intervals across all five chromosomes (Table 1; Figure 2A). This marker set allows mapping at a 10 cM (~2 Mbp) resolution.

To test the efficiency of linkage mapping with the new polymorphic marker set, a Ler line carrying T-DNA constructs overexpressing CONSTANS (CO) and FLOWERING LOCUS C (FLC), as well as a fusion between the promoter of SUPRESSOR of OVEREXPRESSION OF CONSTANS 1 (SOC1) and a beta-glucuronidase (GUS) gene (35S::CO 35S::FLC 1 kb::SOC1:GUS; 14), was subjected to EMS mutagenesis in order to screen for mutations affecting flowering time. An early flowering mutant displaying an elongated hypocotyl when grown in white light was identified in the M₂ generation. After demonstrating that the early flowering mutant phenotype was stably inherited to the M₄ generation, the mutant was back-crossed with the progenitor transgenic Ler parent. F₁ progeny showed wild type flowering time (hereafter referred as normal flowering time), indicating that the mutation was recessive. Recessive inheritance was confirmed by the analysis of 96 F₂ progeny, of which about one quarter showed early flowering. Subsequently, a Col-0 mapping parent was generated by crossing the 35S::CO/35S::FLC/1 kb::SOC1:GUS transgenes from the Ler progenitor to Col-0 four times. The early flowering Ler mutant line was subsequently crossed with the Col-0 mapping parent, and a segregating F₂ population was generated to map the mutation. Six early flowering plants were chosen from the F₂ generation, and one leaf from each plant was bulked together and DNA purified using the protocol of Edwards et al. 9. Similarly 6 plants showing a normal flowering phenotype were identified, their leaves were also bulked together and DNA purified.

DNA fragments were amplified from each of the DNA bulks by PCR using the entire new set of 51 poylmorphic markers. Two markers, ciw3 and F26B6 on chromosome II were identified to be linked to the early flowering mutation as both markers were homozygous for the Ler allele from the early flowering bulk and heterozygous for the normal flowering bulk. The other DNA markers were heterozygous in both bulks indicating that they were not linked to the early flowering mutation, with the exception of three markers that most likely were linked to the loci of the T-DNA carrying the CO, FLC and SOC1 gene constructs (data not shown). Figure 2B shows the linked marker F26B6 and an unlinked marker F14G16 amplified from the Col-0 and Ler parents and the DNA bulks of early and normal flowering F₂ progeny. DNA marker F26B6 was confirmed to be linked to the mutation by PCR amplification of the marker from five early flowering and four normal flowering plants from the F₂ mapping population (Figure 2C).

We then used the Sucrose Prep to rapidly screen 96 early flowering plants from the F₂ population, confirming that the mutation was located between the markers ciw3 and F26B6 (data not shown). Analysis of the annotated DNA sequence for candidate genes within this region revealed PHYB as one of the most likely candidates responsible for the early flowering mutant phenotype. Previously, a phyB mutant has been demonstrated to flower earlier under long day conditions and has an elongated hypocotyl under white light conditions 1516. Therefore, we sequenced the PHYB gene from wild type Ler and our early flowering mutant. A base substitution of cytosine to thymine was detected at position 1660 bp downstream of the PHYB translational start site, resulting in a premature stop codon in the first exon. This nonsense mutation is predicted to result in a truncated protein that is unlikely to be functional as it contains neither the PAS repeat domain nor the histidine kinase related domain essential for the known function of the protein.

To determine if the Sucrose Prep method is suitable for screening of homozygous T-DNA mutants, we screened segregating T₂ progeny from the SALK_098205 line, in which a T-DNA was inserted in exon 3 of the AtWRKY22 gene (At4g01250; Figure 3). In parallel, we isolated RNA from the same plants using a commercial kit. As illustrated in Figure 3, the Sucrose Prep (DNA; upper panel) produced results that were consistent with the observed expression of the gene (cDNA; lower panel) thereby identifying line 3 as a homozygous loss-of-function mutant of AtWRKY22.

The simplicity of the Sucrose Prep also facilitated the screening for T-DNA insertions in larger populations. For example, the identification of double knockouts carrying T-DNA insertions with identical selectable markers is readily feasible using the Sucrose Prep procedure. To illustrate this point, we performed an experiment to create an atwrky46 (At2g46400), atwrky53 (At4g23810) double mutant. These WRKY transcription factors belong to the same sub-group (group III) and show very similar transcription induction kinetics in response to pathogen and elicitor treatments (data not shown). After crossing the T-DNA insertion lines (Figure 4), F₂ progeny were screened by PCR for the loss of gene specific products. Seven candidates were immediately identified as potentially being homozygous for both insertion mutations within 90 plants assayed (Figure 4B). Two of these 7 candidates were subsequently confirmed to be true double knockouts (not shown).

To demonstrate that the Sucrose Prep method is also suitable for facile detection of specific transcripts in small amounts of plant tissue, we tested the expression of SOC1 (At2g45660) and AGC1-10 (At2g26700) genes in leaf samples by RT-PCR amplification of RNA templates prepared by the Sucrose Prep. Since RNA might be destroyed during heating, the extracts from leaf samples were subjected to various heating times of 1 to 5 min prior to RT-PCR amplification. The length of the heating step did not appear to influence the efficiency of RT-PCR amplification since the SOC1 transcript was detected in all samples (Figure 5). Although the amplification of the cDNA product was weaker in comparison to the efficiency obtained with the commercial RNA purification kit, the cDNA product was specific and its amplification was reproducible.

Next, we tested homozygous T-DNA knockout lines for loss of WRKY transcripts. A T-DNA in the AtWRKY36 (At1g69810) gene was localized within the first intron. This insertion resulted in a loss of detectable transcript by RT-PCR analysis compared to wild type (data not shown). RT-PCR analysis of wild-type and homozygous atwrky36 knockout mutant plants was performed by using the Sucrose Prep. Since the T-DNA was located in the first intron, no amplification of cDNA product was expected in the knockout mutant (Figure 6, panel A). Nonetheless, a faint cDNA product was detected by the RT-PCR assay suggesting that the T-DNA insertion was spliced out from a small fraction of primary transcripts.

We also developed an alternative method for isolating nucleic acids from very small sample sizes designated 'Touch-and-Go'. This method eliminates the preparation steps required by the Sucrose Prep, since the extraction of DNA/RNA templates is made simply by capturing plant tissue with a pipette tip which is immediately available for the amplification of nucleic acids by PCR and RT-PCR methods respectively. In practice, leaf tissue was punctured using an RNAse-free 20 μl pipette-tip against a firm surface, i.e. a finger covered with a latex glove, and then the pipette-tip was immediately touched into 50 μl PCR solution mix in prepared reaction wells. Due to the very small amount of leaf material taken by pipette tip puncturing, the number of PCR cycles should be 35 to 40 when using this method. For plant material located in the greenhouse or the field, 10 μl water was first delivered to PCR tubes or plates kept on ice, and 'Touch-and-Go' sampling was performed by touching the water in the tubes or wells with the pipette tip containing the plant tissue. After sampling, 40 μl PCR solution mix was added to the tubes or wells in the laboratory and DNA/RNA was amplified by PCR for 40 cycles using a thermocycler. Figure 6, panel B illustrates that the 'Touch-and-Go' method can be used to PCR amplify DNA products a maximum of 1.5 kb in size. This mini-preparation method was also tested in combination with RT-PCR by monitoring for the loss of detectable expression of AtWRKY70 (At3g56400) in a T-DNA insertion mutant line. The atwrky70 insertion mutant carries the T-DNA insertion in the last exon of the gene. Figure 6, panel B illustrates that no specific cDNA signal was detectable in the homozygous atwrky70 knockout line in comparison to wild type extracts from which both DNA and cDNA products were well amplified. To verify that the observed lower band of the correct predicted cDNA size of is indeed AtWRKY70, this fragment was gel isolated and subsequently sequenced. The sequencing data clearly confirmed that this fragment represents the full-length AtWRKY70 cDNA fragment and that the two known introns were spliced out.

The 'Touch-and-Go' extraction method was also tested in screening for an atwrky40, atwrky18 double mutant. Figure 7 shows a screening of F₂ progeny homozygous for the atwrky18 T-DNA insertion mutation. Whereas a DNA fragment of 644 bp specific for the AtWRKY18 locus was amplified from extracts prepared from wild type Col-0 and homozygous atwrky40 mutant plants, this was not the case when the assay was applied to extracts that were prepared from homozygous atwrky18 lines. By screening 72 segregating F₂ progeny from an atwrky40 × atwrky18 cross, we identified 19 individuals to be homozygous for the atwrky18 mutation (as judged by the absence of the 644 bp PCR product). These results are in agreement with the expected Mendelian segregation ratio (i.e. 18/72; Figure 7)

We isolated DNA from the dicotyledonous crop species tobacco (Nicotiana tabacum) and its close relative N. benthamiana, as well as from the monocotyledonous crop species barley (Hordeum vulgare) using Sucrose Prep and the 'Touch-and-Go' methods to demonstrate their applicability for plants other than A. thaliana. As illustrated in Figure 8 (upper two panels), both methods produced sufficient quality and quantity of DNA that can be amplified using primers for two tobacco genes NtCDPK2 (calcium dependent protein kinase2) and NtRBCS that yield 2 kb and 0.8 kb PCR amplified DNA products respectively. Amplification failed in only one out of eight reactions using the Sucrose Prep while none failed using the 'Touch-and-Go' method. A similar isolation and PCR amplification was performed with barley tissue using four barley specific primer pairs producing varying sizes of amplified DNA fragments (Figure 8; lower two panels and Table 2). The sizes and patterns of the observed amplified DNA fragments are identical to those observed with other conventional DNA isolation methods (personal communication T. Zhao, M. Böhmer, and G. Freymark, MPIZ Köln). All of the primer pairs produced the expected size fragments (1.7, 0.7, 0.6 and 0.4 kb) in PCR analysis using DNA isolated by the Sucrose Prep. Using the Touch-and-Go' method, three primer pairs successfully amplified the expected smaller size fragments (0.7, 0.6 and 0.4 kb) but failed to amplify the largest size fragment of 1.7 kb.

Flow diagrams for the Sucrose Prep and the 'Touch-and-Go' methods for PCR and RT-PCR applications are shown in Figure 9.

Sucrose Solution: 50 mM Tris-HCl pH 7.5, 300 mM NaCl and 300 mM sucrose.

(A) Individual samples: approximately 10 mg of leaf tissue was placed directly in 200 μl Sucrose Solution and ground at room temperature or on ice using a pipette tip or pestle. The samples were then heated to 99–100°C for 10 min and then briefly spun at 2000–6000 g for 5 sec. The samples were placed on ice until PCR. One μl of the supernatant was used for PCR, avoiding debris.

(B) Multiple sample 96-well format: between 10–20 mg leaf tissue from 14 d-old plants was harvested as leaf discs into 96 well plates. Metal balls (3 mm, tungsten carbide beads) were added and shaken in a Retsch MM300 shaker for 10 sec, then 300 μl of Sucrose Solution was added, the plate heated at 99°C for 10 min and placed on ice until use. Following storage at 4°C or -20°C, samples were reheated at 99–100°C for 10 min and then immediately placed on ice.

Leaves were punctured against a firm surface, like a finger covered with a latex glove using a fine pipette tip (TipOne from Starlab GMBH, catalog no. S1111-3000 or S1110-3000). The DNA/RNA on the tip of the pipette was transferred to the pre-prepared PCR solution in the PCR tubes by touching the tip of the pipette to the solution and pipetting up and down a few times. For plants in the field or greenhouse, 10 μl water was aliquoted into PCR tubes or microtiter plates. The tubes/plates were kept on ice while puncturing the leaves with a fine pipette tip against a firm surface and DNA/RNA the tip was transferred into water by pipeting up and down. After returning the samples to the laboratory, 40 μl PCR master mix was added to each well. Thermocycling with the 'Touch-and-Go' method requires 40 cycles of PCR amplification.

The CTAB protocol used was developed by Murray and Thompson 32, modified from Rogers and Bendich 33 and adapted by Rios et al. 2.

A. thaliana plants were grown under a 16 h photoperiod at 20°C in a greenhouse.

Approximately 6,500 35S::CO 35S::FLC 1 kb:SOC1::GUS co-2 seeds were mutagenized by imbibition in 0.3% ethyl methanesulfonate (EMS; Sigma) for 8 to 9 h, followed by washing with 0.1 M Na₂SO₄ and distilled water. The M₁ mutagenized seed was planted into about 260 pools each containing 25 M₁ seeds. About 300 M₂ seed from each pool were sown under long day conditions and scored for flowering time.

Routine PCR: 3 min 94°C, 35–40 cycles of: (30 sec 94°C, 45 sec 55°C, 1 min 72°C), 10 min 72°C, 4°C until analysis. 2.5 μM each gene specific primers, 2.5 mM dNTPs, 5–10 U (0.5–1 μl) Taq polymerase (Invitrogen), 1 × Taq Buffer (commercially supplied). For products larger than 2 kb, 0.5 U of enzyme LA Taq polymerase (TaKaRa) was substituted for Invitrogen Taq polymerase and the PCR protocol from Rios et al. 2 was used.

Qiagen OneStep RT-PCR Kit (catalog no. 210210) was used following the manufacturer's recommendations. For isolation of cDNA, RNA was extracted with RNeasy Plant Mini Kit from Qiagen (catalog no. 74904) following the manufacturer's instructions.

Primers and T-DNA knockout lines are listed in Table 2.

All of the agarose gel pictures are ethidium bromide stained gels and the images are inverted in Adobe Photoshop.